Embed Size (px)
Citation preview
Test Rig Design for Accelerated Life Testing of Chainsaws
A Major Qualifying Project Report
Submitted to the Faculty of the
WORCESTER POLYTECHNIC INSTITUTE
In partial fulfillment of the requirements for the
Degree of Bachelor of Science
by:
Olivia Steen (WPI)
Adam Olsson (KTH)
Jacob Andersson (KTH)
Johan Erkers (KTH)
Marcus Söderberg (KTH)
Oskar Lundkvist (KTH)
Sponsored by:
Joakim Arvby
Husqvarna Petrol Handheld Product Laboratory, Husqvarna AB
Date: 17 May 2018
Professor Holly Ault, Advisor
Worcester Polytechnic Institute
Professor Kjell Andersson, Project Supervisor
Royal Institute of Technology
Professor Stefan Björkland, Project Supervisor
Royal Institute of Technology
This report represents the work of WPI undergraduate students. It has been submitted to the faculty as
evidence of completion of a degree requirement. WPI publishes these reports on its website without
editorial or peer review.
i
Abstract
This project presents a conceptual design for a test rig for accelerated life testing of chainsaws.
Three different subsystems, loading, oil collection, and handles, were created to reflect realistic
operating conditions in the testing environment. The loading subsystem applies forces to the bar
that mimic the normal forces felt by the saw during field operation. The oil collection subsystem
consists of three chambers to collect the oil mist and aid in the increased air purification. Finally,
the handles constrain the saw in a way that mimics the response of a field operator. These
subsystems were then integrated, producing the full concept of the rig. Further development and
integration of these subsystems will be continued, and a prototype will be built in the coming
semester.
ii
Acknowledgements
The authors, Adam Olsson, Anders Karlsson, Johan Erkers, Jacob Andersson, Olivia
Steen, Marcus Söderberg and Oskar Lundkvist would like to thank and acknowledge the
following persons for their help and guidance throughout the project. Their help and guidance
has been crucial in the development of this project. Special thanks goes out to:
Joakim Arvby, Group Manager R&D Global Services at Husqvarna Group
Kjell Andersson, Professor at department of Industrial Engineering and Management,
KTH.
Stefan Björklund, Associate professor at department of Industrial Engineering and
Management, KTH.
Martin Engholm, Testing Engineer at Husqvarna Group
iii
Preface
This project stems from a collaboration between the Royal Institute of Technology (KTH)
and Husqvarna AB for the KTH project course MF2076 during the spring semester of 2018, under
the Industrial Engineering and Management department at KTH. The project fulfills the Major
Qualifying Project for Worcester Polytechnic Institute and a partial fulfillment of the Machine
Design master program for KTH students. The design generated in this report serves as the basis
for the prototype to be completed by KTH students in the fall semester of 2018.
iv
Table of Contents
Abstract .......................................................................................................................................................... i
Acknowledgements ....................................................................................................................................... ii
Preface ......................................................................................................................................................... iii
List of Figures .............................................................................................................................................. vi
1. Introduction ........................................................................................................................................... 1
2. Background ........................................................................................................................................... 1
2.1 Current Husqvarna Testing Laboratory Problems .............................................................................. 2
3. Project Requirements ............................................................................................................................ 3
3.1 Requirements for Loading Engine and Chassis .................................................................................. 3
3.2 Requirements for Oil Systems ............................................................................................................ 4
3.3 Requirements for Restraining the Chainsaw ....................................................................................... 4
4. Concept Generation .............................................................................................................................. 5
4.1 Loading System .................................................................................................................................. 5
4.1.1 Form Based Concepts .................................................................................................................. 5
4.1.2 Chosen Concept ........................................................................................................................... 7
4.2 Oil Collection System ....................................................................................................................... 10
4.2.1 Materials Selection for Covers ................................................................................................... 14
4.2.2 Vacuum/Air System ................................................................................................................... 15
4.2.3 Simulation of Exhaust Gas Flow ............................................................................................... 17
4.2.4 Future Work on Oil System ....................................................................................................... 20
4.3 Handle System .................................................................................................................................. 20
4.3.1 Transmissibility and damping frequency response of hands ..................................................... 21
4.3.2 Handles Design Concepts .......................................................................................................... 21
4.3.3 Clamping mechanism ................................................................................................................. 23
4.3.4 Free body diagram ..................................................................................................................... 26
4.3.5 Dynamic models ........................................................................................................................ 27
4.3.6 Final Handle Concept proposal .................................................................................................. 29
4.3.7 Future Work on Handles Design ................................................................................................ 31
5. Conclusion .......................................................................................................................................... 32
References ................................................................................................................................................... 34
Appendix A: Project Requirements and Specifications .............................................................................. 35
Appendix B: Friction Based Loading Designs ........................................................................................... 37
Appendix C: Oil Collection Cover Material Selection Matrix ................................................................... 38
v
Appendix D: Equations of First Dynamic Model ....................................................................................... 40
vi
List of Figures
Figure 1 Sketch of Cogwheel Form-Based Concept ..................................................................................... 6
Figure 2 Chain and Timing Belt Concept Shown Attached to Bar ............................................................... 6
Figure 3 Depiction of Unmodified Chainsaw Chain Segment ...................................................................... 7
Figure 4 Depiction of Modified Chain Segment ........................................................................................... 8
Figure 5 Loading System CAD Highlighting Pivot Point and Torque Transfer ........................................... 8
Figure 6 CAD of Modified Chain on Timing Belt ........................................................................................ 9
Figure 7 Sprocket Component of Timing Belt Concept ............................................................................. 10
Figure 8 CAD of Three Cover System for Oil and Exhaust Collection. ..................................................... 10
Figure 9 Bar Cover Design ......................................................................................................................... 11
Figure 10 Clutch Cover Design .................................................................................................................. 12
Figure 11 Image of Husqvarna 550 XPG Chainsaw Muffler ..................................................................... 13
Figure 12 Conceptual Design of Exhaust Cover ......................................................................................... 13
Figure 13 Sentry Air Systems Mist Collector Model # SS-300-MIST (SAS Inc, 2016) ............................ 16
Figure 14 COMSOL Geometry of Chainsaw Exhaust Port Deflector ........................................................ 18
Figure 15 COMSOL Set-Up of Exhaust Port Simulation ........................................................................... 19
Figure 16 COMSOL Simulation Results, Streamlines of Exhaust Gases Velocity Fields Exiting Muffler 19
Figure 17 Handles Concept 1, stiff arms and rubber clamps ...................................................................... 22
Figure 18 Handles Concept 2, adjustable dampers and rubber clamps ....................................................... 22
Figure 19 Handles Concept 3, adjustable damper and mass resembling body ........................................... 23
Figure 20 Varying Force Clamping Mechanism Concept .......................................................................... 24
Figure 21Constant Force Clamping mechanism Concept ........................................................................... 25
Figure 22 Free body diagram of the chainsaw in the xy-plane ................................................................... 26
Figure 23 Free body diagram of the chainsaw in the xz-plane ................................................................... 26
Figure 24 First dynamic model ................................................................................................................... 28
Figure 25 Second dynamic model ............................................................................................................... 29
Figure 26 Pugh's evaluation matrix of the handle concepts ........................................................................ 30
Figure 27 Final Concept Design ................................................................................................................. 32
1
1. Introduction
The purpose of this project is to develop and create a test rig for accelerated life testing of
Husqvarna 50cc chainsaws. Accelerated life testing is when products are tested at high speeds and
stresses for extended periods of time to discover failures; these tests are a necessary part of
ensuring the product is safe to use (Intertek Group, n.d.).
Husqvarna currently has test boxes that run chainsaws for many hours to see how the saw
takes the wear. However, the conditions in the test box do not match the conditions in the field, so
the company relies on field data, which is an expensive and lengthy process. The goal of this
project is to design a testing structure that better mimics the field test environment in terms of
loading, lubrication, and vibration. This test rig design can then be implemented in future test
boxes to limit the amount of field testing required.
2. Background
Chainsaw manufacturers must test new products in accordance with ISO standards to verify
the useful lifespan, the time before the first major failure, of the product. The necessary lifespan
of a chainsaw depends on the target customer, with professional saws needing to last at least 400
hours (HVA/KTH Project 2018, 2018). The number of useful operating hours can be determined
through accelerated life tests.
Accelerated life testing is a form of testing wherein a product is subject to increasing levels
of variables encountered in operation. During chainsaw accelerated life tests, the manufacturers
must investigate the crucial elements of the machine’s duty cycle (the start-up, the operational run,
and the run-down) though predefined test cycles (Pesik and Skarolek, 2014). Several methods of
2
accelerated life testing have been done in chainsaw manufacturing, including using a water brake
load and user cutting tests.
In a water brake load, water inside a casing is used to simulate the loading the saw would
sustain while cutting a log. The casing can fill with a variable amount of water to simulate different
loads. The torque of the motor is then derived from the shear forces in the water measured by a
load cell inside the casing (Feldkamp and Tedesco, 2003). The drawback to this type of
measurement device is that it can only be used in above zero temperature or else the water will
freeze (Joesfsson and Henningsson, 2015).
User cutting tests provide the most accurate results, as they are more representative of how
the saw will be used in the field. However, these tests are harder to perform as they require a human
operator and a large amount of timber. Thus, simulated load tests using water brakes are performed
in testing laboratories. These require less man-hours and resources, but do not accurately reflect
the engine and chassis loading experienced by the saw in the field (HVA/KTH Project 2018, 2018).
2.1 Current Husqvarna Testing Laboratory Problems
Husqvarna's testing laboratory currently houses 30 chainsaw testing cells. These cells run
tests with both water brakes using ISO standard test cycle TC-H17-02, and running the engine
with a bar but no chain using ISO standard test cycle TC-H17-01 (HVA/KTH Project 2018, 2018).
However, the damages that appear on the saw while cutting timber are not reflected in the test box
simulations due to the lack of realistic loading. Therefore, to get a clear idea of the effectiveness
of the chainsaw design Husqvarna must rely on field tests, where data is collected from operators
using the new saw in non-test environments.
3
Field testing is a great way to get input on the developed product, however it can be very
time consuming, costly and provides feedback regarding the product too slowly. To generate more
immediate feedback, Husqvarna wants to upgrade their testing facilities to include test rigs that
better simulate reality.
3. Project Requirements
Before the work process of the project could begin, Husqvarna presented several areas that
should be addressed in the creation of a test rig to simulate realistic engine and chassis loading.
The first focus areas are the chassis and engine; these components must be loaded to reflect the
normal cutting forces experienced by the bar during operation. The handles are another important
part of the saw. A method for restraining the saw in a way that realistically models the resilience
and frequency response of a user holding the handles. A system should also be implemented to
collect the used oil, preventing the oil cloud that is currently created in the test cell. Specifics for
each of these areas are laid out below. A complete list of requirement specifications and tasks can
be found in Appendix A: Project Requirements and Specifications.
There have been several limitations we have kept in mind throughout the concept
generations. The solutions must fit in the supplied test box dimensions and utilize a fully assembled
chainsaw. There can be no increase in fire hazard, and the solution should be able to be easily
implemented and maintained by testing technicians.
3.1 Requirements for Loading Engine and Chassis
The test rigs currently in use at Husqvarna have no way to simulate loading of the bar. A
water brake attached to the front end of the bar brakes the chain during a set of test cycles. Although
the water brake brakes the chain, it lacks the ability to load the bar in a realistic way due to its
4
placement. The loading system must brake the chain and load the bar in a manner that reflect the
normal cutting forces to better represent the fatigue and wear experienced during field testing. To
achieve this a number of concept ideas were generated and evaluated.
The saw and bar should withstand 400 hours while the chain is replaced multiple times
during the saw’s lifetime. But even so, the chain fatigue should be taken into account. The fatigue
of the bar will both be evaluated through calculations and simulations in the test rig. The load on
the bar in its normal direction will affect the joints where the bar is connected to the main body of
the saw. To simulate the normal cutting forces, a change in height and an angle between the bar
and horizontal plane are needed. If possible the test rig should be constructed in a way so that the
readings represent the actual wear on the engine caused by usage of the saw.
3.2 Requirements for Oil Systems
During field testing most of the excess oil on the bar and chain is absorbed by the cutting
material. In the current test box, the saw is run with the standard amount of oil and no cutting
material. These conditions are leading to an excess of oil, resulting in runoff and oil mist not
reflective of realistic operating conditions. Reducing the cloud of oil mist is a crucial part of
creating more realistic cutting conditions.
3.3 Requirements for Restraining the Chainsaw
The main requirement and purpose of the handle subsystem is to mimic a gloved operator
using the chainsaw during normal operation. The requirements for the entire test rig system were
evaluated and simplified as to which may interfere and depend on the handles subsystem.
5
4. Concept Generation
To start generating possible concepts, the group used a brainstorming technique called 6-
3-5 method. This method consists of six people, or seven in this case, generating three ideas each
during five minutes. The ideas are then passed on to the next member for refinement and feedback.
The group used a compressed and modified version where each paper was passed on so that each
paper was evaluated by two more people other than the original creator. The session then restarted
until the group were satisfied with the result. Looking at Husqvarna's requirements and the ideas
generated from the 6-3-5, the group decided to divide into the three subgroups described below:
loading, oil and handles.
4.1 Loading System
The loading ideas generated in the brainstorming sessions: friction-based concepts, where
a friction material is used to brake the chain, and form-based concepts, where geometries
complimentary to the chain are used to brake the chain. After evaluation it was decided that the
friction based methods would generate too much heat, causing the box temperature to exceed
Husqvarna’s requirements. The discarded friction based concepts can be found in Appendix B:
Friction Based Loading Designs. Below are short descriptions of the possible form-based solutions to
fulfill the given requirements.
4.1.1 Form Based Concepts
Cogwheel. The initial idea for a form-based load solution was to have a sprocket that the
chain could hook on to, loading the engine in a realistic way. This solution would also make it
possible to provide a load on the bar as shown in Figure 1. This method was discarded since the
loading scenario of the engine was not that realistic since the solution only providing a point
6
contact between the chain and the sprocket. However, this idea was further developed into a
concept involving a chain and timing belt.
Figure 1 Sketch of Cogwheel Form-Based Concept
Chain and Timing Belt. This solution was developed based on the first form-based idea.
Unlike the first solution, the load is distributed over a larger area. This gives a more realistic load
case for the engine and the bar load scenarios. Using a chain or timing belt, the idea is to hook into
the chainsaw chain as shown in Figure 2. This is the chosen concept to be further developed and
will be described in more detail.
Figure 2 Chain and Timing Belt Concept Shown Attached to Bar
7
4.1.2 Chosen Concept
Out of the above evaluated concepts it was decided that the most realistic one to move on
with would be the Chain/timing belt solution. This solution requires a modified chain to transfer
the torque from the chain to the braking device. This is a basic modification that only changes the
assembly order of links. The original chain consists of four different kind of links and is shown in
Figure 3 Depiction of Unmodified Chainsaw Chain Segment. The names of the links are as follows: 1)
Knife 2) Standard Link 3) Drive Link with Tip 4) Drive Link
Figure 3 Depiction of Unmodified Chainsaw Chain Segment. The names of the links are as follows: 1) Knife 2) Standard Link 3)
Drive Link with Tip 4) Drive Link
The original chain is not suitable to transfer the torque, since its design doesn't allow the
chosen solution to interact with the chain in a proper way. The proposal for modification is shown
in Figure 4. This chain consists of only link types 2 and 4 to get the drive link shape on both sides
of the chain. This shape is convenient for the torque transfer.
8
Figure 4 Depiction of Modified Chain Segment
To be able to transfer the torque evenly when the chainsaw oscillates from vibrations, the
connection between the vertical force (mounting to the ground) is done by a bridge with its pivot
point above the transfer plane, see Figure 5. This whole upper part is connected to the lower part
with, still to be chosen, a vertical load/displacement applier. The connection between the chainsaw
chain and brake is still to be evaluated further, but two concepts for this presented below.
Figure 5 Loading System CAD Highlighting Pivot Point and Torque Transfer
9
The modified chain is placed on a double timing belt setup, as seen in Figure 6. The timing
belts are connected to each other with small steel rods that the chain hooks into, and two sprockets
are connected to two shafts, where the brake is mounted. The chain is always in contact with the
timing belt rods in order to reduce wear and unnecessary shocks from impact. To cancel out the
inertial effects the chainsaw experiences from this the design, an accelerator is connected to one
of the shafts as well. This regulates the speed of the timing belts when no load is applied to match
the speed of the chain.
Figure 6 CAD of Modified Chain on Timing Belt
One of the advantages of using timing belts is that they are somewhat flexible and thus
spread the load between the points of contact. Timing belts are also suitable for the speed of the
chainsaw. One disadvantage is the potential impact of the oil mist, if the timing belts starts to slide
the braking would not work. The sprockets, as seen in Figure 7, are paired with a distance in
between for the shark fins to have clearance.
10
Figure 7 Sprocket Component of Timing Belt Concept
4.2 Oil Collection System
The main purpose of the oil problem solving concept is to increase purification rate in the
test-box and increase the oil collection rate. This is achieved by using the concept illustrated in
Figure 8, which consists of several different sections. To prevent the mist cloud from causing over-
lubrication and contaminating the intake air we have separated the system into its three critical
areas. Each area is isolated by a housing that collects the oil for recirculation, preventing the mist
from contaminating the test box air. These three critical areas addressed are the bar/chain, clutch
cover, and exhaust fumes.
Figure 8 CAD of Three Cover System for Oil and Exhaust Collection. 1) Bar Cover 2) Clutch Cover 3) Exhaust Cover
11
The Bar Cover. A box encompassing the bar and chain will collect the excess oil coming
off the groove. Figure 9 is a concept drawing of the separation chamber. The bottom of this
chamber has a sloped surface with suction to draw oil mist to the bottom of the chamber, into the
filtration system, and allow it to be collected and recirculated. The dimensions of this chamber
must also be able to house the loading device. Another condition that must be addressed is how
this separation chamber would affect the heat transfer of the system. The material must have a
sufficient heat transfer coefficient to allow for proper heat conduction and keep the bar and ambient
temperature at acceptable levels.
Figure 9 Bar Cover Design
The Clutch Cover. Much of the oil that is not captured in the previous collection system is
disposed of as the chain enters the clutch cover, creating an oil pool and oil mist. To prevent this
mist, a second separation chamber that covers the area of the clutch is integrated into the system.
This chamber must surround all sides of this area to be able to capture all the mist. Similar to the
bar and chain chamber, the caught oil will be directed to a reservoir for recirculation. The cover
design must keep in mind the parameters set out by the handle design and must not cause
12
overheating in the system. This design is illustrated below in Figure 10. A similar vacuum system
is attached to the clutch cover area.
Figure 10 Clutch Cover Design
The Exhaust Cover. The last critical area addressed by the three-chamber system is the
exhaust gases that contaminate the intake. The exhaust gases that exit the motor through the
muffler (shown in Figure 11) contain a significant proportion of oil. The current system has no
way to direct these gases away from the intake, causing dirty air to re-enter the motor. In this
solution the exhaust gases are separated from the intake portion of the motor through a hose
attached directly over the muffler, filtering the exhaust gases out from the test-box. The vacuum
air system must be powerful enough to redirect most of the gases exiting the port into the vacuum
tube, as the nozzle should not be directly touching the chainsaw to limit the influence the tubing
has on the vibrations of the system. The nozzle should have a wide enough opening to capture
most of the exhaust gases.
13
Figure 11 Image of Husqvarna 550 XPG Chainsaw Muffler
This nozzle over the muffler will nullify the exhaust gases, eliminating interference
between exhaust gases and intake. A limitation for this exhaust cover is it cannot be made of metal,
due to the generation of sparks in the muffler area if a metal nozzle were used. The conceptual
design of the exhaust cover is illustrated in Figure 12.
Figure 12 Conceptual Design of Exhaust Cover
14
4.2.1 Materials Selection for Covers
The materials must also be selected for each of the three chambers. This should be done
through a selection matrix of the different criteria. The material must:
Withstand greater than the ambient temperature in the test box
Be durable enough to survive the long-term vibrations of the saw system
Be easy for technicians to remove and replace
Not interfere with the temperature of the system
Be easily obtainable and workable
Not increase fire hazards
Has a high durability when in contact with oil/corrosive environment
A matrix comprised of ten different materials was created to compare the properties of potential
potential materials. The properties were found using the material database CES Edupack. The table
table containing the material properties can be found in of oil and added cooling.
15
Appendix C: Oil Collection Cover Material Selection Matrix.
One of the main requirements of this design is that it cannot increase the fire hazard.
This criterion ruled out any material in our matrix that is flammable. The nickel-chromium alloy
and stainless steel are then removed due to their comparatively high cost combined with their
average machinability and moldability. The remaining three materials were low carbon
steel, aluminum alloys and copper. The workability of low carbon steel is lower than that of the
aluminum and copper, however the cost of the steel is much cheaper. Comparing the aluminum
and copper, both have great workability, but copper is more expensive so to save on costs that
material is not selected. Thus, aluminum has been selected as the potential material to create the
bar and clutch covers.
This selection should be verified through thermodynamic analysis once the
dimensions have been agreed upon to prove that the box systems do not cause significant heating
of the saw and ambient air. The options for material for the exhaust cover need further
development, as discussed in Section 4.2.4 Future Work on Oil System
4.2.2 Vacuum/Air System
The vacuum system attached to the bar and chain cover must provide enough suction to
prevent the pooling of excess oil and collect as much of the excess as possible before the chain
enters the clutch cover. Most of the oil needed to be collected in this area is in the form of oil mist,
created by the flinging of oil off the bar and chain system. This mist is sucked by the air system
and then filtered before the air is returned to the test chamber. Thus, the system must be powerful
enough to move the oil particles in the given box area, as well as the filter must be able to
accommodate the size of the oil mist particles, which generally range from 0.1 microns to 0.3
microns (Kittelsona, 1998).
16
The vacuum system attached to the clutch cover works in a smaller volume than the bar
cover, but the filter requirements and particle size are the same. The exhaust cover vacuum system
should accommodate exhaust gases and, like the previous critical area systems, be able to filter the
oil mist particles.
Vacuum systems used to extract oil mist in similar settings were researched to find suitable
systems. Oil mist extraction units, such as the ones shown from Sentry Air Systems in Figure 13,
are used in machine shops where lubricated tools generate oil mist.
Figure 13 Sentry Air Systems Mist Collector Model # SS-300-MIST (SAS Inc, 2016)
This air system can handle an air volume of up to 100 CFM (cubic feet per minute) and
can be supplied with a filter that is up to 99.99% effective on particles with diameters down to .12
microns in size (SAS Inc, 2016). A system such as this could be connected to the tubes attached
to the bar cover, clutch cover, and exhaust cover.
Once the specifications of the vacuum system are known, the dimensions of the tubing
attached to each cover can be more accurately determined. The pressure generated by the vacuum
17
system is determined by the area of the tube opening. The variance in pressure between the vacuum
tube and the surrounding area should exert a large enough force to collect the mist particles.
A rough estimation of these pressures can be found using the Bernoulli equation. This can
be used because the area the air is flowing from a larger area (the area of the box) through a smaller
area (the area of the tube) with an increasing velocity as it is sucked into the tube due to the pressure
difference created by the vacuum system. Bernoulli’s equation is listed below.
𝑃1 +1
2𝜌𝑣1
2 + 𝜌𝑔ℎ1 = 𝑃2 +1
2𝜌𝑣2
2 + 𝜌𝑔ℎ2 (1)
For all covers, the particles are assumed to start suspended in still air, thus v1 is zero. The
variable h1 depends on the geometry of the cover, with the max h1 being used in the estimate. This
is the distance of the area to be acted upon in reference to the pipe opening, h2, which is zero. In
these calculations, P1 is the pressure of the box, assumed to be atmospheric pressure, or Patm, while
P2 is the pressure inside the vacuum system. This P2 is unknown. To solve for P2, or Pvac, the
equation is rearranged to find Equation 2:
𝑃𝑣𝑎𝑐 = 𝑃𝑎𝑡𝑚 +1
2𝜌𝑣2
2 − 𝜌𝑔ℎ1 (2)
The variable v2 is dependent on the geometries of the tube openings in the individual
chambers, as it is defined by Equation 3 in which the flowrate, defined by the flowrate of the air
system, is divided by the cross sectional area of the tube opening.
𝑣2 = 𝑓𝑙𝑜𝑤𝑟𝑎𝑡𝑒/𝐴𝑐,𝑡𝑢𝑏𝑒 (3)
Once Pvac is calculated, the force acting on the oil mist particles can be found as in Equation
4, wherein force is defined as the pressure acting on the particle (Patm) multiplied by the cross
sectional area of the particle. If this force is greater than the force of gravity acting on the particle
18
then the tube-vacuum system will be sufficient to move the particles into the filter. These equations
can be used in future work to refine the geometries of the tube openings and needed vacuum
systems.
𝐹𝑎𝑝𝑝𝑙𝑖𝑒𝑑 = 𝑃𝑣𝑎𝑐 ∗ 𝐴𝑐,𝑡𝑢𝑏𝑒 (4)
4.2.3 Simulation of Exhaust Gas Flow
Before the design of the exhaust cover, the flow of the exhaust gases had to be understood.
To understand the movement of the exhaust, the muffler was simulated in the program "COMSOL,
2017". A flow simulation was used to understand the dispersion of exhaust gases out of the muffler.
After looking at the velocities and flow directions of the fumes the shape of the exhaust nozzle
best suited for capturing them was designed.
The initial velocity of the exhaust gases as they exit into the muffler had to be calculated
to act as input parameters into the COMSOL simulation. The ingoing velocity of the exhaust gases
entering the muffler were calculated in Equation 5, where dispersion is the stroke volume of the
engine in cubic meters and Aout is the area of the outtake of the engine measured in square meters.
𝑉𝑜𝑢𝑡 = 𝑉𝑖𝑛 = 𝑑𝑖𝑠𝑝𝑒𝑟𝑠𝑖𝑜𝑛 ∗𝑠𝑝𝑒𝑒𝑑 𝑜𝑓 𝑒𝑛𝑔𝑖𝑛𝑒
2∗ 𝐴𝑜𝑢𝑡 = 5.371 [
𝑚
𝑠] (5)
The geometry of the deflector on the chainsaw was measured, and then modeled in Solid
Edge and converted into a COMSOL geometry, which is illustrated in Figure 14. This geometry
was then analyzed by also creating an ingoing cylinder to act as ingoing exhaust gases to the
deflector, and by creating a larger box act as an empty room, in attempt to track the velocity flow
of the model, seen in Figure 15.
19
Figure 14 COMSOL Geometry of Chainsaw Exhaust Port Deflector
Figure 15 COMSOL Set-Up of Exhaust Port Simulation
20
Figure 16 COMSOL Simulation Results, Streamlines of Exhaust Gases Velocity Fields Exiting Muffler
Figure 16 displays the velocity behavior of the exhaust gases from the muffler. This result
indicated that most of the exhaust gases are spread upwards and towards the sides of room when
exiting the deflector. The result also indicates that the exhaust cover displayed in Figure 16 would
capture most of the exhaust gases leaving the deflector, but this will have to be further adjusted
and verified during the autumn semester.
4.2.4 Future Work on Oil System
The bar cover will need further adjustments next semester, mainly due to the further
development of the loading concept. These require more information from the loading concept and
also calculations concerning the temperature of the system. The temperature will be calculated
with analytical thermodynamic equations or/and investigated with programs such as COMSOL.
The clutch cover will need dimensioning and thermodynamic analysis. Another interesting
factor to investigate would be whether or not the clutch cover solution affects the vibration of the
21
chainsaw system, because that could create issues for the handles concept. This would also need
to be investigated for the bar cover and the exhaust cover.
Once the dimensions and vacuum systems are selected, calculations similar to those
discussed in Section 4.2.2 Vacuum/Air Systemwill be performed to understand the necessary tube
openings and vacuum properties.
The exhaust cover needs verification and selection of a material. It was noted by the project
sponsor that the material for the exhaust cover cannot be metal, as that causes sparks and increases
fire hazards. Potential materials to solve this problem will need to be further investigated.
4.3 Handle System
To begin the process of designing a handle system, a dynamic model of the system was
made so the frequency response of the human hand could be better understood. In accomplishing
this goal, extensive research had to be performed regarding Husqvarna's chainsaw range, the
magnitude of the forces acting on the bar, the potential damping of plastic materials at specified
operating frequencies, and vibration measurement methods. Mathematical models were then used
to simulate the behavior of the chainsaw during load. Physical vibration measurements will later
be performed to verify the model.
4.3.1 Transmissibility and damping frequency response of hands
While running the chainsaw, an operator is exposed to vibrations, measured in m/s^2 with
a specified amplitude and frequency determined by the chainsaw construction and running mode,
and the operator is exposed to force impulses when the cutting chain initially touches the log as
well as the vibrations caused by the engine. These accelerations are transferred to the user via the
points of contact at the two handles and the transferred amount is determined by the transmission
22
ratio T. The transmission ratio is described as the difference in vibration amplitude of two surfaces
in contact or as the difference in displacement as a function of time of two surfaces in contact
represented in the equation below (Wallin, 2014).
𝑇 =𝐹𝑜𝑢𝑡
𝐹𝑖𝑛=
𝑥𝑜𝑢𝑡
𝑥𝑖𝑛 (6)
4.3.2 Handle Design Concepts
The first concept has the chainsaw mounted to two static arms with rubber clamps, shown
in Figure 17 below. This concepts focus is to only resemble the damping characteristics of human
hands. It is easy to implement in the test cell and is good in a maintenance point of view.
Figure 17 Handles Concept 1, stiff arms and rubber clamps
The second concept is a modified version of the first one. In addition to the rubber clamps,
it is hanging in the test cell with adjustable dampers. The purpose of these dampers is to simulate
the resistance a user would give to the forces that are applied by the loading group who will add
forces on the bar. The dampers can be adjusted to work with different chainsaw models if needed.
Concept 2 is shown in Figure 18 below. Other benefits to adjustable dampers include the possibility
of simulating user fatigue wherein the user gets weary and cannot provide the same reaction force
during a full load cycle.
23
Figure 18 Handles Concept 2, adjustable dampers and rubber clamps
The third concept has a mass attached to the chainsaw handles resembling the hands and
arms of a user, also including an adjustable damper so it can be tuned to match the behavior of the
human body. Figure 19 below shows the third concept. This concept has an inertia that the other
two concepts do not have which might be possible to tune.
Figure 19 Handles Concept 3, adjustable damper and mass resembling body
4.3.3 Clamping mechanism
The main purpose of the handles is to resemble a user holding onto the handles. The length
of the handle should be approximately 100 mm for hands according to studies and 13 mm extra
24
since gloves are used when operating the chainsaw (Lindqvist, 1998). The maximum allowed
pressure in the clamping surface is 20N/cm^2 and the clamping mechanism must fit tightly around
the handle.
Simulating user fatigue by varying the grip force would be interesting because there are
indications that the transmissibility coefficient T may vary as a function of grip force and
frequency according to R. Gurram, S. Rakheja & Gerard J. Gouw (1994). The concept for the
clamping mechanism was developed where the clamping force can be varied using a compact
cylinder. The clamping mechanism concept is shown in Figure 20 below, the picture shows the
clamping mechanism for the top handle.
Figure 20 Varying Force Clamping Mechanism Concept
The concept consists of a rubber cylinder in two halves, pushed together by a compact
cylinder. The clamping force should be possible to vary between zero and the force that
corresponds to the maximum allowed pressure in the clamping surface.
25
The alternative to having a concept with adjustable grip force would be to clamp with an
unvaried, constant force. One way to do it is by using the same method that is used on bicycle
wheels, the quick release technique. A concept proposal using this technique is shown in Figure 21
below.
Figure 21Constant Force Clamping mechanism Concept
The selection of material for the clamping concepts is a complicated matter due to the
required properties specified below. The material needs to have good durability to oil as well as
good fatigue and damping properties. Ideally there would be no oil on the clamping mechanism,
but the oil collection subsystem cannot be assumed to 100% eliminate oil in the test cell and must
therefore be resistant to the used chain lubrication.
A lot of rubbers have bad durability against oil. There are some synthetic rubbers that are
durable to oil, but natural rubber is often better when it comes to most factors. An alternative to
having rubber would be PE-foam, due to its good durability against oil. Its relative damping ratio
and fatigue strength are, however, lower compared to natural rubber. Apart from the
26
aforementioned properties, the most crucial material property to study when choosing a material
is how the mechanical properties such as damping, and spring coefficient of chosen rubber may
be affected by the variance in temperature.
4.3.4 Free body diagram
Models of the system were created to design and dimension the components for the
concepts. A free body diagram is performed for the XY plane and the XZ which is shown in Figure
22 and Figure 23 below. The loading group will apply a set of forces on the blade in point C. There
are reaction forces in point A, the midpoint of the position where the user is holding the front
handle and point B, the midpoint of the position the user is holding the rear handle. The dimensions
shown in the figure were measured in the CAD-model of the chainsaw provided by the sponsor.
Figure 22 Free body diagram of the chainsaw in the xy-plane
27
Figure 23 Free body diagram of the chainsaw in the xz-plane
4.3.5 Dynamic models
A dynamic model is preferred to dimension the springs and dampers that will be attached
to the handles. It will also simulate how all parts of the system behave when running at different
chainsaw engine speeds and loaded externally. The model has the potential to be used by
Husqvarna for other purposes such as ergonomic studies or optimization of geometry.
The model will be loaded with vibrations from the motor as well as external forces on the
chainsaw bar. The dynamic model is divided into several steps due to complexity. The first model
is of the chainsaw treated as one body and the second one takes the internal spring and dampers
into account. The following assumptions are made for the two models to simplify the design
process.
Assumptions and simplifications:
The chainsaw body itself is assumed to be rigid compared to the springs and
dampers attached.
28
The forces are only acting in one plane. This simplifies the calculations and can be
derived, assuming that forces acting in the z-direction are taken up by the log if
there are any at all.
The inertial effects from the movement of the chain along the bar are neglected.
Forces and displacements are only working in the same direction as the dampers
and springs, which means a damper in x-direction does not take any forces in the
y-direction.
Point forces and masses are assumed for all interactions between parts and rigid
walls.
A numerical method is used to calculate the results in Matlab.
The first model is shown in Figure 24 below. It has 3 degrees of freedom, it can translate in
x and y direction and rotate around the center of mass and is attached in the cell with a spring and
damper in each direction for both rear and front handle. Inputs needed by this model are the inertia
for the chainsaw, the position of the center of mass, the vibrations from the motor, the loading
forces Fy & Fx as well as the mass of the chainsaw. A Matlab code is written to calculate the
position, velocity and acceleration of the handles with a defined set of dampers and springs.
29
Figure 24 First dynamic model
Equations of the system are obtained and using them and the inputs on the external forces
the results are calculated using ODE45 in Matlab. The equations are shown in Appendix D:
Equations of First Dynamic Model. The final step of the dynamic model is to sum the accelerations
in the handles and compare them to the values measured during next semester as well as the values
Husqvarna’s own measurements.
Further development for the dynamic model gives the opportunity to take the internal
springs and dampers of the chainsaw into account of the observed behavior. There are several
springs and dampers between the handles and the motor and blade which need to be taken into
account to resemble the actual chainsaw properly. The second model is shown in Figure 25 and it
will be modeled in Matlab during next semester. It has 6 degrees of freedom, x- and y-direction
for both masses as well as rotations around the center of masses.
Both of the models will be compared and verified against the measured results as well as
the data provided by Husqvarna to determine the validity as well as the usability by the models.
30
Figure 25 Second dynamic model
4.3.6 Final Handle Concept proposal
To compare the concepts with each other and get an overview a Pugh's evaluation matrix
is made. In the matrix different requirements on the handles subsystem is weighted on a scale 1-5
and concept one is used as a reference. The other two concepts are then given a plus or a minus if
it is better or worse than that concept when it comes to meeting the requirements. This gives an
overview of which concept has which edge, some of the requirements have been assumed. Hand
resemblance is set to be the most important requirement and it can only be verified by
measurements. The matrix is shown in Figure 26 below.
31
Figure 26 Pugh's evaluation matrix of the handle concepts
The conclusion from this matrix is that the concepts are similar when it comes to meeting
the requirements that can be evaluated on a conceptual stage. All the concepts can for example use
the same clamping mechanism and should have equally low effect on the other two subsystems.
The first concept seems to meet these requirements best, followed by the second one.
However, it is clear that the most important requirements such as hand resemblance need to be
evaluated based on measurement results. The dynamic model might give some hints on what kind
of design is better but measurements are needed to draw a proper conclusion and select concept.
The third concept will be not be used as it seems difficult to succeed with adding a tuned mass.
When it comes to clamping mechanism concept it is undecided which of the concepts that
may provide the best solution and therefore measurements are needed to show the effects on the
vibration levels from the variance in clamping force.
32
4.3.7 Future Work on Handles Design
With the free-body diagrams and main ideas for the dynamic model derived, the future
work on the model would be creating a finished MatLab code. This model may give insights as to
which conceptual design best models human hands. However, to understand which concept is able
to mimic these vibrations, vibration tests must first be completed. The physical testing methods
will need to be drawn up and the tests performed. The data from the results will identify which
dampers and springs systems to use.
33
5. Conclusion
The conceptual design for a test rig for accelerated life testing of chainsaws that better reflects
realistic operating conditions was addressed thought the generation of conceptual designs in three
different subgroups. A design for a device to apply a load mimicking that of the normal forces felt
by the saw is applied to the bar, as seen in Section 4.1.2 Chosen Concept. Section 4.2 Oil Collection
System shows the three chambers making up the oil collection system. These chambers collect the
oil mist generated by the excess oil as well as separate the exhaust to aid in an increased
purification rate in the test cell. Finally, the handles seen in Section 4.3.3 Clamping mechanism were
applied to constrain the saw in a way that mimics the response of a field operator.
Figure 27 Final Concept Design
34
All three concepts are combined in the figure above, Figure 27, showing the final integrated
system. Further work must be done on to understand the interactions between the three concepts.
Vibrations are one potential negative interaction. It is important that the loading system and the
oil system do not have a significant effect on the vibrations of the chainsaw because that could
interfere with the vibrations results extracted from the handles as well as alter the performance of
the chainsaw.
The most critical interaction to address is between bar loading and bar cover. The bar cover
should be large enough to encapsulate the loading solution and collect the oil that lubricates the
different loading surfaces. Once the loading device is finalized, the interactions between both of
these components can be evaluated.
With most of the conceptual work completed, the remaining work to be completed in KTH
next autumn will be able to focus on detailed dimensioning and understanding the fully integrated
system. After the finalization of these designs, the full prototype will be built and tested to evaluate
the effectiveness of these concepts.
35
References
B. Lindqvist. Verktygsergonomi. Atlas Copco, Stockholm, 1998.
Feldkamp, K., & Tedesco, A. (2003, December 11). Water Brake Dyno. Retrieved from
http://courses.me.berkeley.edu/ME102B/Past_Proj/f03/Proj11/discuss.htm
Gurram, R., Rakheja, S., & Gouw, G. J. (1994). Vibration transmission characteristics of the human hand-
arm and gloves. International Journal of Industrial Ergonomics,13(3), 217-234.
doi:10.1016/0169-8141(94)90069-8
HVA/KTH Project 2018 [Interview by J. Arvby]. (2018, March). At Husqvarna Petrol Handheld Product
Laboratory.
Intertek Group. (n.d.). HALT Testing & HASS Testing. Retrieved from
http://www.intertek.com/performance-testing/halt-and-hass/?gclid=CjwKCAiA-
9rTBRBNEiwAt0Znwy4zoIVlmIx-
glOcjeMocWeS5i8LIxgUxhYpka5uTsljfrG7C5XxzRoCRAUQAvD_BwE
Joesfsson, E., & Henningsson, H. (2015). A Study of Small Engine Tests. Tekniska Hogskolan Hogskolan
I JonKoping. Retrieved from http://www.diva-
portal.se/smash/get/diva2:860427/FULLTEXT01.pdf
Kittelsona, D. B. (1998). Engines and nanoparticles: A review. Journal of Aerosol Science,29(5), 6th ser.,
575-588. Retrieved from https://doi.org/10.1016/S0021-8502(97)10037-4.
Pesik, I., & Skarolek, A. (2014, May 1). Dynamic model of a chainsaw. At Technical University of
Liberee. Print.
(SAS) Sentry Air Systems, Inc. (2016). CNC Machine Mist Collection. Retrieved from
https://www.sentryair.com/200mistsentry.htm
Wallin, H. (2014). Ljud och Vibrationer. Universitetsservice, Stockholm.
36
Appendix A: Project Requirements and Specifications
The following table contains the project specifications and tasks as defined by our
sponsor.
37
38
Appendix B: Friction Based Loading Designs
Disc Brake. One of the initial ideas where to brake the chain using a standardized disc brake
solution. This concept would be divided in two sections, one that can be seen in the below drawing
that brakes the chain, the latter to apply a load to the bar. The loading concept are described in
general under other headings in this chapter and therefore not discussed further here. The braking
pads would brake the chain from the sides as can be seen from the top view in the conceptual
drawing below, where the clamps would be fixated using a rack that would be fixated on the floor.
The clamps would be dimensioned according to the necessary force needed but using standard
components as much as possible.
Figure 28 Disc Brake Concept Sketch
Vertical Loading. A second friction-based concept was evaluated, where loading and
braking would be applied using a single vertical movement. The friction material would be applied
to a broader plate that would cover the chain’s width and brake it while the plate would be moved
upwards to apply a load to the bar. The chain would have to be modified to replicate the
39
engagement between a regular saw chain and a log to avoid the standard chain getting stuck in the
friction material.
Figure 29 Vertical Loading Concept Sketch
V-Shaped Braking. Below a concept, based on the other two friction concepts, will be
described. The thought was to combine the advantage of the disc brakes braking solution with the
loading of the bar in a more realistic way than for the "Vertical loading friction concept". The
thought was to better be able to control the friction applied when loading the bar with the "normal
force", the angle of the "V" would be optimized to replicate the situation of the chain engaging a
log. One major problem here was to find a suitable friction material sturdy enough to avoid wear,
creating guidelines in the material and therefore alter the angle in an unwanted way. All the above
friction concepts including the present one would require the absence of oil and added cooling.
40
Appendix C: Oil Collection Cover Material Selection Matrix
Flamability
Working
Temperature
Durability
to Oil Machinability Moldability
Cost
(SEK/kg)
Toxic
Rating
Stainless Steel
non-
flammable -272C to 750C Excellent 2 to 3 3 to 4 48-50
non-
toxic
Low Carbon Steel
non-
flammable -68C to 350C Excellent 3 to 4 3 5-7
non-
toxic
Cast Al Alloys
non-
flammable
-273C to 130-
220C Excellent 4 to 5 4 to 5 16-18
non-
toxic
Ni-Cr Alloy
non-
flammable -272 to 900C Excellent 3 3 119-131
non-
toxic
Wood (plywood)
Highly
flammable -100C to 120C Acceptable 5 3 to 4 5
non-
toxic
CFRP (carbon
fiber)
Slow
burning -123C to 140C Excellent 1 to 3 4 to 5 321-356
non-
toxic
PLA
Highly
flammable -20C to 45C Acceptable 4 to 5 4 to 5 22-30
non-
toxic
PET
Highly
flammable -123C to 67C Excellent 3 to 4 4 to 5 15-17
non-
toxic
PMMA
Highly
flammable -123C to 56C Excellent 3 to 4 4 to 5 24-25
non-
toxic
Copper
non-
flammable -273C to 300C Excellent 4 to 5 4 to 5 43-51
non-
toxic
41
Appendix D: Equations of First Dynamic Model
